Affiliation
American Association of Variable Star Observers (AAVSO)
Wed, 05/06/2020 - 20:33
I have two CCD cameras: an SBIG ST-7E-T and an Orion Starshoot G4. The ST-7E-T is a NABG type and has a FWC of 100,000 electrons while the Starshoot G4 has a FWC of only 50,000 electrons. However, they have approximately the same dynamic range. The ST-7E-T has a read noise of 15 electrons while the Starshoot G4 has a read noise of only 7 electrons. Also, I did test the linearity of the Starshoot G4 for the CCD 1 course and it's linear up to 50,000 ADUs. My question do NABG-type CCD cameras have other atributes beside the larger FWC that make them a worthwhile investment?
Hi Tim,
Remind me of what FWC stands for?
The SBIG camera uses the KAF0400E NABG sensor; the G4 uses the Sony ICX829ALA ABG sensor. They have roughly the same format (number of pixels and pixel size), but the Sony sensor will be much more blue sensitive, and has higher QE in general (77%). So asking whether ABG or NABG is better, based on two different sensors from different vendors, is not quite valid.
In general, NABG sensors tend to be more sensitive with bigger full well depth for the same pixel size. The ABG process uses up some of the possible pixel sensitive area, which is probably the reason for the smaller well depth. In addition, adding a blooming gate makes response nonlinear near the full well, so most vendors limit the digitized range to about half full well to get decent linearity (which further reduces the effective full well depth). From the same vendor, the ABG sensor tends to be cheaper than the NABG cousin; I think this is because of volume (more ABG sensors made).
That said, it is really hard to compare one sensor vs. another. Comparing CCD vs. CMOS gets even harder, especially when you add in front-illuminated vs. back illuminated versions of each. All of these digital imagers work; just find the one that fits your requirements.
Arne
I was seriously considering using the Starshoot over the ST-7 except two things have made him hesitant to do so
In 2017 I made an unfiltered observation of the HADS V593 Lyr with a Orion Starshooter G3 Mono, just to test it. It behaved very well. Here is the lightcurve. So it can be done.
I saw your post on CloudyNights. Excellent work, BTW.
I do have some questions. What about the constinency of your temperature control. My Starshoot G4 could fluxuate by several degrees, though I just remembered some of the issue involved taking short subs like bias and flat frames.
CCD temperature for flat frames need only match CCD temperature for the (bias and) darks frames applied to those flats; it's not so important to match flat CCD temperatures to the working light frames. I think you'll see that this simplifies things considerably.
CCD temperature fluctuations during light-frame acquisition are harder. First, I would run some tests to see whether the fluctuations really matter for your purposes. If not, problem solved.
If they do matter, some possibilities are: (1) do something to stabilize the CCD temperature, even if only approximately (a bit of insulation? cheap thermostat and dew heater?); (2) take separate darks at a high-mid temperature and other darks at a low-mid temperature and do your best to match darks to light frames (really tedious and error-prone unless automated); or (3) take darks at some sort of average temperature and just use those (if temp range is maybe 5-6 C or less). I would prefer (1) to (2) to (3).